We live in exciting times. There are a few key technologies that are all still in their infancy, all of which I expect to see dramatic innovation and growth in, and all of which "play off" one another in different ways. This is stacking on top of continued innovation around cloud, infrastructure automation, DevSecOps, serverless, and many other technologies which are still relatively early in their adoption curve. The next-gen innovations I see playing a pivotal role in society are:
- Artificial Intelligence - in particular, the subset of AI that is Machine Learning, and in particular Deep Learning; the incredible success of convolution neural networks in processing images for use cases like Autonomous Driving Systems is an example of evolution, but mountains of research are being done trying to apply many techniques in novel ways - some in academia, some in the private sector. We are rife with companies using AI to teach computers to learn to understand sound, or compose music, or translate between languages in real time or call and make appointments for you (speech is, after all, the API of choice for so many human interactions)
- Edge Connectivity and Computing - or if you prefer the more holistic continuum, fog computing, which will become a system for connecting events, people, devices, and computing power at very high speed in diverse locations. There's no beating the speed of light (yet), but design patterns in event-based and flow-based architectures are being advanced and promulgated among software engineers, and advances in hardware, software, frameworks, design patterns, and the increased distribution of fog computing power to address these use cases will unlock new tiers of applications that allow more devices, newer use cases, and indeed entire new categories of value. A popular example conceptualizes being able to play AAA video games on a phone or tablet, utilizing game console hardware at a nearby edge, with only the interface transmitted to the user. With latency and throughput for proximate users that is better than what most experience on wifi today, the possibilities are big. Industrial applications in healthcare, manufacturing, and so on are also relatively low-hanging fruit.
- Augmented Reality - Virtual Reality is fun and it also has a place - especially in training and therapy - but Augmented Reality is, in many ways, a way to finally deliver what Steve Jobs described as the bicycle for our minds. The challenge is that only so many activities benefit from a computer. It's a lot - across so many areas of science, finance, and more they are a transformative tool. Doubly so when you encompass how they form the Internet. But at the end of the day, the interfaces of the finger swipe and the keyboard and mouse have limits - and the output on a screen and audio are even more extreme limits. Augmented reality allows computers to overlay incredibly rich information onto our senses that enhances what we perceive in the real world. Initial applications will be fast references - like instructional videos and manuals which float in the air, virtual walls which indicate safety barriers, and more. Over time, however, augmented reality can allow for much more complicated training and reference interactions, and combined with AI, can then analyze the performance of tasks in the real world, not only tuning the playback of the instructional material but monitoring for defects in the performance. All this leads to humans being able to work much faster and smarter. You can bet I wish this existed the last time I had to disassemble half my oven to change the fan.
- Robotics - The final pillar of transformation in the real world is robotics. You might think from Boston Dynamic's YouTube channel that the primary use of robotics is to delight viewers with robot antics, but the reality is that in world with rapidly evolving AI and edge bandwidth and processing, you have the fundamental pieces to combine with the evolution of robotics hardware and software to enable them to replace huge swaths of thankless human effort.
These don't operate in a vacuum. A host of other technologies are required in their own way in order to have a true "fourth industrial revolution" - for example, if we imagine a massive explosion in the use of computing power, large deployments of robotics (including things like drones), it's easy to imagine a huge uptick in the use of power. Solar or other renewable sources then becomes more critical to not create an environmental catastrophe. Because so many of the technologies have the potential to greatly impact the efficiency, scale, and safety of manufacturing, it is easy to imagine that these collectively can ramp production and consumption across the globe.
nVidia recently announced their extension of the omniverse to a large new set of users (or builders, if you prefer). It's fascinating in some ways how many areas of technology nVidia is simultaneously enabling:
- Their GPUs are used for training in a variety of AI use cases - these are GPUs which may even lack the instruction set for rendering graphics the way a "classic GPU" did
- Their GPUs are also used for understanding and rendering 3d spaces, whether purely virtual or based on reality - naturally for gaming and VR, the classic pure GPU use cases, but also now for augmented reality for the real world
- Together those technologies can also create augmented reality scenes based on the real world, with understanding of real motion, and down to comparative modeling. Creating the so-called "digital twin" of a real-world environment can then allow applications that enhance safety, efficiency, or introduce robotic assistance
- Their data center networking via their Mellanox group also helps enable, in dense computing environments, high performance computing to enable more efficient and effective use of computing resources
It's amazing how much growth is already tied to this revolution when we have barely scratched the surface.
Meanwhile, the early use of drones has some amazing applications. Why send someone to precariously wander across your roof after a hailstorm to inspect it for damage, when a drone can survey and record the roof, and then utilize AI to detect damage? Better to be faster, more accurate, and safer.
Why douse farmland with pesticides if you can use robotics, AI, and edge connectivity to laser down 100,000 weeds per hour? This is far from the only application in farming. There's a ton of work being done in smart farming, and I'm continuously reminded that vast, cheap energy from solar - augmented by smart mining, smart manufacturing, and more - can take use cases that previously seemed absurd and make them real.
It may seem like all this points to a lot more consumption - and it certainly does on the energy front. While the materials side of the equation is certainly impacted by increased demand, it will also be impacted by enhanced sharing economies. You know that tool you had trouble finding the other week? The one you only use every year or so? Wouldn't it be easier to just click a button on an app and have it dropped off for an hour by a small drone, and then picked back up when you indicated you were done? I'd be a lot happier with that, and the app-based catalog, than yet another search through my basement.
As I wrote about on Forbes, a key challenge in all this advancement is people. We need a lot more engineers. Because so much of technology stacks on other technology, we need people tackling these new problem spaces; we need insatiable curiosity, we need a passion for creation - and we need to realize that there are huge areas of development with profound growth which are progressing more slowly because they are held back in other areas.
These technologies tie together the "Faster and Faster" economy as the fourth industrial revolution progresses. There are profound environment and social challenges around the bend with the change this brings. But like Steve Job's "bicycle for the mind" mapped to the entire "real world", we are also seeing the power of technology move beyond the screen to touch the real world in profound ways and that has only scratched the surface.
